Comparative analysis of YOLO Models for Real-time Personal Protective Equipment Detection(PPE)

Linia Mathew; Dr. Vijayalakshmi N

Comparative analysis of YOLO Models for Real-time Personal Protective Equipment Detection(PPE)

Authors: Linia Mathew, Dr. Vijayalakshmi N

Unique Paper ID: 174861
Volume: 11
Issue: 11
PageNo: 1370-1385

Keywords: Computer Vision Deep Learning Image Processing Object Detection Occupational Safety PPE Detection Real-time Monitoring Workplace Safety YOLOv5 YOLOv8 YOLOv11

Abstract:
Personal Protective Equipment (PPE) detection is an essential aspect of maintaining workplace safety, especially in dangerous settings like building construction sites. This paper provides a comparative study of three deep learning object detection YOLOv11, YOLOv8 and YOLOv5 for PPE detection. The research utilizes a dataset from Roboflow that is composed of labelled images of protective equipment like helmets, vests and gloves. All models are compared using relevant performance indicators including accuracy, precision, recall, mean Average Precision (mAP), and inference speed. All the results clearly show that YOLOv11 performs better than its counterparts in terms of detection accuracy as well as computation efficiency, thereby making it an effective option for real-time applications of PPE monitoring. A detailed analysis is also performed based on dataset distribution, inference time, and computation needs to study the effectiveness of these models for real-time scenarios. The research identifies the need for choosing an efficient object detection model to improve labor safety and minimize occupational risks.

Download article

email to a friend

Cite This Article

ISSN: 2349-6002
Volume: 11
Issue: 11
PageNo: 1370-1385

Comparative analysis of YOLO Models for Real-time Personal Protective Equipment Detection(PPE)

Available:https://ijirt.org/Article?manuscript=174861

Impact Factor
8.01 (Year 2024)

UGC Approved
Journal no 47859

Join Our IPN

IJIRT Partner Network

Submit your research paper and those of your network (friends, colleagues, or peers) through your IPN account, and receive 800 INR for each paper that gets published.

Join Now